Secondary Logo

Journal Logo


Flumazenil expedites recovery from sevoflurane/remifentanil anaesthesia when administered to healthy unpremedicated patients

Karakosta, Agathi; Andreotti, Beatrice; Chapsa, Chara; Pouliou, Aikaterini; Anastasiou, Emmanouil

Author Information
European Journal of Anaesthesiology: November 2010 - Volume 27 - Issue 11 - p 955-959
doi: 10.1097/EJA.0b013e3283398ef9
  • Free



Anaesthetics exert their effects largely through affecting the neuron-to-neuron signalling in the central nervous system (CNS).1 The effects on the molecules underlying chemical synaptic transmission, such as γ-aminobutyric acid (GABA), are well decribed.2 Compelling evidence from experimental research and in-vitro studies implicate the inhibitory action of γ-aminobutyric acid type A (GABA-A) in the anaesthetic mechanism of action.3 The GABA-A receptor is the major inhibitory ion channel in the mammalian brain. Benzodiazepines enhance the effect of GABA by interacting with the GABA-A receptor as allosteric modulators and acting at the benzodiazepine binding site; a distinct binding site at the GABA-A receptor.4

Endozepines (endogenous benzodiazepines) are naturally occurring small organic molecules devoid of peptidic bonds and halogens. They are allosteric modulators of the GABA receptor acting at the benzodiazepine binding site. Endozepines are present in physiologically significant amounts in the brain.5 They have also been isolated in urine,6 cerebrospinal fluid7 and breast milk8 of women who had not received benzodiazepines.

Flumazenil, a 1,4-imidazobenzodiazepine derivative, reverses the hypnotic/sedative action of benzodiazepines on the GABA-A receptors. Flumazenil demonstrates high affinity for the benzodiazepine recognition site located in the GABA-A receptor and has little or no agonist activity. It seems to act as an analeptic agent when administered to patients with hepatic encephalopathy,9,10 whereas a recent study showed that it significantly increases the bispectral index (BIS) value and allows earlier emergence from anaesthesia when administered to healthy unpremedicated patients during propofol/remifentanil anaesthesia.11

Incremental intravenous bolus injections of flumazenil 0.1–0.3 mg are considered most effective in the treatment of benzodiazepine overdose, whereas 1 mg is reported as the most frequent maximal dose in humans.12 The efficacy of a large bolus dose of flumazenil has already been documented.13,14 However, to our knowledge, the analeptic role of flumazenil when administered in small clinical doses in unpremedicated patients who were not taking benzodiazepines or other sedative/hypnotic medications has not been fully investigated.

The primary objective of the present study was to investigate the hypothesis that a single bolus injection of flumazenil 0.3 mg, given intravenously (i.v.) at the end of the surgical procedure, expedites the emergence from sevoflurane/remifentanil anaesthesia. Secondary objectives were to investigate safety under these treatment conditions.



After approval by the institutional Human Ethics Committee and having individual written informed consent, 24 patients undergoing elective surgery and requiring general anaesthesia were randomized by sealed envelope allocation to receive either a single dose of 0.3 mg flumazenil (n = 15) or placebo (n = 10) i.v. at the end of the surgical procedure.

Exclusion criteria included weight disorders (BMI <18 or >30 kg m−2) and treatment with sedative/hypnotic agents or any agent that could interfere with GABA metabolism or with the GABA receptor function. In addition, clinical conditions that may affect the level of consciousness, such as strokes, stupor or dementia, were excluded. Inclusion criteria referred to clinically healthy individuals under 70 years of age, ASA physical status I or II and individuals who were scheduled for surgery with an estimated duration of less than 2 h and an estimated blood loss of less than 1 l intraoperatively.

Study protocol

Baseline measurements of vital signs were recorded preoperatively. All patients were administered a small dose of fentanyl i.v. (2 μg kg−1 of body weight) prior to intubation in order to reduce the haemodynamic reaction to intubation. Anaesthesia was induced with an 8% concentration of sevoflurane. Muscle relaxation was accomplished either by administration of succinylocholine or by nondepolarizing muscle relaxants (cisatracurium or rocuronium) according to the recommended adult dosage for intubation. Arterial blood pressure and heart rate were monitored before and after intubation in order to assess the haemodynamic response.

Cardiorespiratory monitoring consisted of electrocardiography, automatic oscillometric cuff blood pressure, side-stream capnography and finger plethysmography-based oximetry. These measurements and the data concerning the anaesthetic agents were continually recorded throughout the surgery. The lungs were mechanically ventilated with 40% O2 (inspired concentration) in air enriched with sevoflurane. The end-tidal concentration of sevoflurane was set to 1.5% and the ventilation was adjusted to maintain an end-tidal concentration of carbon dioxide (CO2) of 30–40 mmHg.

Remifentanil (0.1–0.3 μg kg−1 min−1) continuous infusion was used for maintenance, whereas fentanyl was also administered intraoperatively in bolus i.v. doses in order to maintain a pain-free state. The level of neuromuscular blockade was evaluated using the train-of-four (TOF) trial. Muscle relaxant agent was readministered in order to maintain T1 at 25%. All patients were warmed using hotline air blankets to maintain a core temperature higher than 36°C (measured with an oesophageal thermometer). In order to meet the patients' fluid requirements, crystalloid solutions were used. Blood loss was estimated from suction bottles and swab/drape weighing and was replaced by 6% hydroxyethyl starch 130/0.4. Ocular microtremor (OMT) monitoring in conjunction with monitoring of physiological assessments indicating the level of sympathetic stimulation was used in order to establish the depth of anaesthesia. OMT was measured by using the closed-eye piezoelectric technique at specific time points (before/after induction of anaesthesia, before/after intubation, before discontinuation of volatile anaesthetic and at full communication). OMT frequencies were used in order to predict movement in response to a stimulus.

Thirty minutes prior to the expected completion of the surgical procedure, the remifentanil infusion was kept constant at 1.5 μg kg−1 min−1. Discontinuation occurred 2 min before the end of the procedure, which was marked as the placement of the final skin stitch. Reversal of neuromuscular blockade (TOF 0.9) was accomplished either automatically or by the administration of specialized antagonists (neostigmine) or selective binding agents (sugammadex) in accordance with suggested dosing guides.

Fourteen patients were administered 0.3 mg flumazenil (flumazenil group) at the end of the surgical procedure. The volatile anaesthetic was discontinued immediately after the flumazenil administration without prior reduction followed by an increase in O2 flow and patient hyperventilation. Ten patients constituting the control group received a similar volume (3 ml) of normal saline (0.9% w/v NaCl/H2O). All investigators and study staff, including the nurses, were blinded to the identity of the injected solution. Both solutions were prepared by the research pharmacy team. There was no difference in the physical appearance of the study drug or placebo.

After discontinuation of anaesthesia, time to spontaneous respiration, eye opening on verbal command, extubation and achievement of communication by date of birth recollection were recorded. Extubation was performed after the following criteria were met: exhibition of spontaneous respiration with satisfactory frequency and tidal volume, recovery of muscle strength and protective airway reflexes. Muscle strength recovery assessment was based on the patient's ability to lift his/her head or to squeeze the anaesthesiologist's hand on command.

The 2 h follow-up period for the safety analysis took place in the postanaesthesia care unit (PACU). All patients were closely followed by the same blinded physician. Any abnormal occurrence was noted and treated accordingly. Patients' discharge and transfer to the ward followed the routine PACU regulations.

Statistical analysis

The study was powered at 80% using a two-tail significance test at the 5% level of significance in order to detect a minimum mean difference of 3.7 min in obtaining full communication (final outcome variable) between the two study groups (9.8 min; SD 3.2 min and 13.5 min; SD 3.1 min for the flumazenil and the control group, respectively). A ratio of 1.5 between the study groups, in favour of the flumazenil group, was also predetermined. This required 15 patients in the flumazenil group and 10 patients in the control group. The minimum important differences and SDs for sample size calculation were based on published literature.11

A total of 25 patients were initially enrolled and randomized to receive flumazenil (n = 15) or placebo (n = 10). However, data from 24 patients were included in the subsequent analysis. One patient was withdrawn from the trial after randomization due to protocol violation. Therefore, 14 and 10 patients constituting the flumazenil and the control group, respectively, met the inclusion criteria.

Data were analysed using STATA 10.1 (Stata Corp., College Station, Texas, USA). The nonparametric Wilcoxon–Mann–Whitney U-test was used to compare patient values between the two groups. Baseline data were analysed using descriptive statistics and values were expressed as mean ± SD. Values of the outcome variables were expressed as the median and interquartile range (IQR). Statistical significance was established at 5% (P < 0.05).


Twenty-four patients, ranging in age from 25 to 70 years, participated in the study. All patients received general anaesthesia for diverse surgical procedures. Ten patients underwent general surgery (five patients in the control and five in the flumazenil group), three patients underwent orthopaedic surgery (one in the control and two in the flumazenil group), two patients underwent plastic surgery (one in the control and one in the flumazenil group), three patients underwent ENT surgery (one in the control and two in the flumazenil group) and six patients underwent gynaecological surgery (two in the control and four in the flumazenil group). Demographic characteristics and baseline data measurements of both groups are shown in Table 1. It appears that the patients in both groups differed only in age, with the patients in the flumazenil group being older than those in the control group.

Table 1
Table 1:
Patient demographics, baseline data, duration of surgical procedure and surgical blood loss

Induction using a volatile anaesthetic was characterized by cardiovascular stability. However, four patients who were administered succinylcholine for optimal intubation conditions (one in the control and three in the flumazenil group) exhibited significant sinus bradycardia during induction and were treated with intravenous atropine. All four patients received sugammadex for neuromuscular blockade reversal at the end of the surgery.

Reversal of neuromuscular blockade was accomplished automatically in one patient in the control and two in the flumazenil group. Neostigmine and concomitant administration of atropine was used for neuromuscular blockade reversal in three patients in the control and three in the flumazenil group. Sugammadex was injected in six patients in the control and nine patients in the flumazenil group. There was no difference in the distribution of overall use of atropine in the two groups (42.8% in the flumazenil group and 40% in the control group). The OMT frequencies at discontinuation of the volatile anaesthetic were similar between the two groups. No movement in response to painful stimulus was recorded intraoperatively in either group.

Time until emergence from anaesthesia was recorded in both study groups. A statistically significant difference was recorded in all recovery parameters when the two groups were compared (Table 2), indicating faster emergence from anaesthesia in the flumazenil group. Specifically, the median time to spontaneous respiration was significantly shorter (P = 0.0003) in the flumazenil group than in the control group [2.5 min (2.0–3.0) vs. 7.0 min (6.8–8.3)], the median time to eye opening on verbal command was significantly shorter (P = 0.0012) in the flumazenil group than in the control group [3.4 min (3.0–4.0) vs. 8.1 min (6.9–10.2)], the median time to extubation was significantly shorter (P = 0.0015) in the flumazenil group than in the control group [4.0 min (3.0–5.0) vs. 9.0 min (7.0–10.8)] and the median time to date of birth recollection was significantly shorter (P = 0.0012) in the flumazenil group than in the control group [4.7 min (4.0–5.0) vs. 10.3 min (8.0–12.0)].

Table 2
Table 2:
Recovery parameters

We encountered no serious cardiovascular events or other adverse reactions throughout the 2 h observational period in the PACU.


Anaesthesia is rarely accomplished by using a single anaesthetic factor. The combination of numerous anaesthetic agents accounts for an unpredicted recovery time. In the present study, induction and maintenance of anaesthesia was achieved using a volatile agent. Inhalation anaesthetics exhibit a more predictable ‘behaviour’ during emergence from anaesthesia than intravenous anaesthetics. This is because intravenous anaesthetics exhibit cumulative actions in cases of dose augmentation and prolonged administration. However, inhalant as well as intravenous anaesthetic agents are implicated in GABA receptor function.15

Volatile anaesthetics enhance synaptic inhibition through GABA in the central nervous system (CNS).16–19 The precise mechanism by which sevoflurane produces loss of consciousness is unknown. It has been demonstrated that sevoflurane and GABA agonists (muscimol, baclofen) act on different domains of the GABA receptor complex.16,17 It has also been demonstrated that sevoflurane activates at least two distinct domains of the GABA-A receptor, increasing the GABA-ergic transmission by enhancing the affinity of GABA for the GABA receptor and inducing a picrotoxin-mediated antagonism of the GABA-A receptor.16,17

In the present study, the analeptic role of flumazenil when administered at the end of the surgical procedure was our main interest. Assessment of the depth of anaesthesia intraoperatively was achieved by both OMT monitoring and the level of sympathetic stimulation. When compared with BIS, OMT presents greater sensitivity (85 vs. 76% for BIS) and greater specificity (94 vs. 69% for BIS) during anaesthesia with sevoflurane and with the end-tidal concentration fixed at 1–2%.20

Through competitive inhibition, flumazenil reverses the hypnotic/sedative effects of benzodiazepines at the benzodiazepine binding site on the GABA receptors. Although sevoflurane is not thought to bind to the benzodiazepine site on the GABA receptor, it seems possible that flumazenil could interact in some way with the GABA receptor to reverse its effect, justifying its role as a general analeptic factor.

Another hypothesis that has been tested in order to justify the analeptic role of flumazenil proposes flumazenil inhibition of the GABA–benzodiazepine receptor complex of endogenous benzodiazepine-like ligands (endozepines) in patients who had not received benzodiazepines.11 Naturally occurring endozepines, or endogenous benzodiazepine receptor ligands, that did not result from environmental contamination with pharmaceutical benzodiazepines have been isolated in different mammalian tissues.11,21,22 Specifically, endogenous benzodiazepines were detected in potentially physiological concentrations in human cerebrospinal fluid, serum, plasma, urine, ultrafiltrates of patients who had not received benzodiazepines and breast milk of healthy, newly delivered women who were not taking benzodiazepines.5–8,23,24 Benzodiazepines of natural origin (NBZDs) have been isolated in medicinal plants and foods.25 A synthetic pathway for the production of NBZDs has not yet been found, but it has been suggested that microorganisms may synthesize molecules with benzodiazepine-like structures.25

Endozepines seem to complicate the role of flumazenil as a general analeptic factor. According to a recent meta-analysis of double-blinded randomized case–control studies, administration of flumazenil induces clinical and electroencephalographic improvement of hepatic encephalopathy in patients with cirrhosis9 and results in a more favourable prognosis.10 Moreover, significant amounts of endozepines have been detected in the brain of patients with hepatic encephalopathy not previously exposed to benzodiazepine medication.26 The authors suggest the presence of nonbenzodiazepine substances (possibly neurosteroids) with positive allosteric modulatory properties at the GABA-A receptor complex in brain in hepatic encephalopathy.26 Furthermore, in a recent prospective randomized clinical trial, administration of flumazenil intraoperatively during propofol/remifentanil anaesthesia in healthy unpremedicated patients not previously exposed to any benzodiazepine medication significantly increased BIS value.11 It has also been demonstrated that flumazenil improves recovery of high cortical and neuromotor functions following halothane, enflurane and isoflurane anaesthesia, reduces shivering and improves the overall quality of emergence.27

References to convulsions and ventricular tachycardia precipitated by flumazenil have been reported in the literature. These are reports of flumazenil administration to comatose patients who had previously received large doses of bezodiazepines.28,29 However, there are no similar reports of patients who had not received benzodiazepines. In the present study, we encountered no cardiovascular events after flumazenil administration.

In conclusion, flumazenil administration during sevoflurane/remifentanil anaesthesia in healthy unpremedicated patients allows earlier emergence from anaesthesia and expedites recovery. This may indicate that flumazenil administration could act as a useful tool in the reversal of exogenous and endogenous benzodiazepines that seem to play an important role during general anaesthesia. However, a large, adequately powered study is needed to carefully examine the safety and efficacy of flumazenil in order to establish its analeptic role in unpremedicated patients at recovery from anaesthesia.


1 Dilger JP. The effects of general anaesthetics on ligand-gated ion channels. Br J Anaesth 2002; 89:41–51.
2 Masaki E, Kawamura M, Kato F. Attenuation of gap-junction-mediated signaling facilitated anesthetic effect of sevoflurane in the central nervous system of rats. Anesth Analg 2004; 98:647–652.
3 Johnston GAR. GABAA receptor pharmacology. Pharmacol Ther 1996; 69:173–198.
4 Sigel E. Mapping of the benzodiazepine recognition site on GABAA receptors. Curr Top Med Chem 2002; 2:833–839.
5 Rothstein JD, Garland W, Puia G, et al. Purification and characterization of naturally occurring benzodiazepine receptor ligands in rat and human brain. J Neurochem 1992; 58:2102–2115.
6 Beaumont K, Cheung AK, Geller ML, Fanestil DD. Inhibitors of peripheral type benzodiazepine receptors present in human urine and plasma ultrafiltrates. Life Sci 1983; 33:1375–1384.
7 Deckert J, Kuhn W, Przuntek H. Endogenous benzodiazepine ligands in human cerebrospinal fluid. Peptides 1984; 5:641–644.
8 Dencker SJ, Johansson G, Milsom I. Quantification of naturally occurring benzodiazepine-like substances in human breast milk. Psychopharmacology (Berl) 1992; 107:69–72.
9 Goulenok C, Bernard B, Cadranel JF, et al. Flumazenil versus placebo in hepatic encephalopathy in patients with cirrhosis: a meta-analysis. Aliment Pharmacol Ther 2002; 16:361–372.
10 Als-Nielsen B, Gluud LL, Gluud C. Benzodiazepine receptor antagonists for hepatic encephalopathy. Cochrane Database Syst Rev 2004:CD002798.
11 Dahaba AA, Bornemann H, Rehak PH, et al. Effect of flumazenil on bispectral index monitoring in unpremedicated patients. Anesthesiology 2009; 110:1036–1040.
12 Weinbroum AA, Flaishon R, Sorkine P, et al. A risk–benefit assessment of flumazenil in the management of benzodiazepine overdose. Drug Saf 1997; 17:181–196.
13 Geller E, Weinbrum A, Schiff B, et al. The effect of flumazenil on the process of recovery from halothane anaesthesia. Eur J Anaesthesiol 1988; 2:151–153.
14 Weinbroum A, Halpern P, Geller E. The use of flumazenil in the management of acute drug poisoning: a review. Intensive Care Med 1991; 17:32–38.
15 Salmi E, Kaisti KK, Metsähonkala L, et al. Sevoflurane and propofol increase 11C-flumazenil binding to gamma-aminobutyric acid A receptors in humans. Anesth Analg 2004; 99:1420–1426.
16 Hirota K, Roth SH. Sevoflurane modulates both GABAA and GABAB receptors in area CA1 of rat hippocampus. Br J Anaesth 1997; 78:60–65.
17 Hapfelmeier G, Schneck H, Kochs E. Sevoflurane potentiates and blocks GABA-induced currents through recombinant α1β2γ2 GABAA receptors: implications for an enhanced GABAergic transmission. Eur J Anaesthesiol 2001; 18:377–383.
18 Franks NP, Lieb WR. Molecular and cellular mechanisms of general anaesthesia. Nature 1994; 367:607–614.
19 Mihic SJ, McQuilkin SJ, Eger EI 2nd, et al. Potentiation of gamma-aminobutyric acid type A receptor-mediated chloride currents by novel halogenated compounds correlates with their abilities to induce general anesthesia. Mol Pharmacol 1994; 46:851–857.
20 Kevin LG, Cunningham AJ, Bolger C. Comparison of ocular microtremor and bispectral index during sevoflurane anaesthesia. Br J Anaesth 2002; 89:551–555.
21 Medina JH, Pena C, Piva M, et al. Presence of benzodiazepine-like molecules in mammalian brain and milk. Biochem Biophys Res Commun 1988; 165:547–553.
22 Klotz U. ‘Natural’ benzodiazepines in man (letter). Lancet 1990; 335:922.
23 Duthel JM, Constant H, Vallon JJ, et al. Quantitation by gas chromatography with selected-ion monitoring mass spectrometry of ‘natural’ diazepam, N-desmethyldiazepam and oxazepam in normal human serum. J Chromatogr 1992; 579:85–91.
24 Wildmann J, Niemann J, Matthaei H. Endogenous benzodiazepine receptor agonist in human and mammalian plasma. J Neural Transm 1986; 66:151–160.
25 Baraldi M, Avallone R, Corsi L, et al. Natural endogenous ligands for benzodiazepine receptors in hepatic encephalopathy. Metab Brain Dis 2009; 24:81–93.
26 Ahboucha S, Pomier-Layrargues G, Butterworth RF. Increased brain concentrations of endogenous (nonbenzodiazepine) GABA-A receptor ligands in human hepatic encephalopathy. Metab Brain Dis 2004; 19:241–251.
27 Weinbroum AA, Geller E. Flumazenil improves cognitive and neuromotor emergence and attenuates shivering after halothane-, enflurane- and isoflurane-based anesthesia. Can J Anaesth 2001; 48:963–972.
28 Marchant B, Wray R, Leach A, Nama M. Flumazenil causing convulsions and ventricular tachycardia. BMJ 1989; 299:860.
29 Short TG, Maling T, Galletly DC. Ventricular arrhythmia precipitated by flumazenil. BMJ 1988; 296:1070–1071.

benzodiazepines; endozepines; flumazenil; recovery; sevoflurane

© 2010 European Society of Anaesthesiology